home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!news.claremont.edu!nntp-server.caltech.edu!brandt
- From: brandt@cco.caltech.edu (William N. Brandt)
- Newsgroups: sci.physics
- Subject: Schrodinger hydrogen atom partition function divergence
- Date: 1 Jan 1993 15:41:50 GMT
- Organization: California Institute of Technology, Pasadena
- Lines: 38
- Message-ID: <1i1onuINNgh@gap.caltech.edu>
- NNTP-Posting-Host: sandman.caltech.edu
- Summary: Schrodinger hydrogen atom partition function divergence
-
- I have a question about the partition function of the Schrodingert
- hydrogen atom. The energy eigenvalues for such an atom go like
- 13.6 ev * (1/n^2) so if we compute Z, the partition function,
-
- Z = Sum[g_n * exp(-E_n/(k*T))]
-
- (g_n is the degeneracy of the nth state, k is Boltzman's constant,
- and T is tempertature) we notice that E_n tends to zero so the
- exponential tends to 1 so we get an explosively diverging series
- since the g_n's tend to increase.
-
- Since Z is not a QM observable this looks at first sight not to be a
- problem. However, since the population of a given level goes as
- 1/Z (and this is an observable) this would seem to indicate that
- the Schrodinger hydrogen atom cannot contain an electron.
-
- I know astronomers resolve this problem by lopping off the partition
- function sum when the bohr radius of the nth state is approximately
- the mean atomic spacing, since EM effects would prohibit higher
- states. However, is there a fundamental resolution to this
- apparent problem? Or is it not really a problem and why? I notice
- that the rigid rotator and harmonic oscillator partition functions
- nicely converge - are there other examples of systems with divergent
- partition functions (I assume most atoms will have similar problems,
- for example).
-
- Does the Dirac equation perhaps remedy this situation, since
- eventually 'tickling' of high n states by QM vacuum fluctuations
- will make very high n states impossible? Or is this not the case
- and why?
-
- I would be grateful if you could send email as well as posting.
-
- Thank you, Niel Brandt
-
- P.S. Any references to this potential problem would also be appreciated.
- P.P.S. Thank to all for answering my previous Feynman graph question.
-
-